A Signature of Roaming Dynamics in the Thermal Decomposition of

Oct 3, 2014 - Pyrolysis of ethyl nitrite, CH3CH2ONO, is studied in a Chen type flash pyrolysis reactor at temperatures of 1000–1800 K. The branching...
0 downloads 0 Views 1MB Size
Download PDF
Letter pubs.acs.org/JPCL

A Signature of Roaming Dynamics in the Thermal Decomposition of Ethyl Nitrite: Chirped-Pulse Rotational Spectroscopy and Kinetic Modeling Kirill Prozument,*,† Yury V. Suleimanov,‡ Beat Buesser,‡,§ James M. Oldham,∥ William H. Green,‡ Arthur G. Suits,∥ and Robert W. Field† †

Department of Chemistry and ‡Department of Chemical Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, United States § IBM Research, Smarter Cities Technology Centre, Dublin 15, Ireland ∥ Department of Chemistry, Wayne State University, Detroit, Michigan 48202, United States S Supporting Information *

ABSTRACT: Chirped-pulse (CP) Fourier transform rotational spectroscopy is uniquely suited for near-universal quantitative detection and structural characterization of mixtures that contain multiple molecular and radical species. In this work, we employ CP spectroscopy to measure product branching and extract information about the reaction mechanism, guided by kinetic modeling. Pyrolysis of ethyl nitrite, CH3CH2ONO, is studied in a Chen type flash pyrolysis reactor at temperatures of 1000−1800 K. The branching between HNO, CH2O, and CH3CHO products is measured and compared to the kinetic models generated by the Reaction Mechanism Generator software. We find that roaming CH3CH2ONO → CH3CHO + HNO plays an important role in the thermal decomposition of ethyl nitrite, with its rate, at 1000 K, comparable to that of the radical elimination channel CH3CH2ONO → CH3CH2O + NO. HNO is a signature of roaming in this system. SECTION: Kinetics and Dynamics

B

and anti- conformations of cyclopropane carboxaldehyderate.2 CP spectroscopy of nascent photolysis products in multiple vibrational states was demonstrated by Field and co-workers, and the approach was proposed for studying transition state(s) properties.12 Recently, in a series of CP experiments, flash pyrolysis was characterized for rotational and vibrational temperatures of species exiting the reactor, and quantitative detection of multiple pyrolysis reaction products was demonstrated.14 The roaming mechanism,20−26 is a recently discovered pathway in chemical reactions, which is proving to be a ubiquitous phenomenon observed in unimolecular laserinitiated,20,27−29 shock tube,30−32 and even bimolecular33 chemistry. In unimolecular decomposition following thermal or laser excitation, the energized precursor molecule (i) may proceed over a barrier, through a conventional tight transition state (TS) on its potential energy surface (PES) to form molecular products; the barriers arise as old chemical bonds are broken and new bonds are formed. Alternatively (ii), a simple bond fission (SBF) reaction may lead to formation of two radicals; since no new bonds are formed, the reaction proceeds

roadband chirped-pulse (CP) Fourier transform rotational spectroscopy, invented by Brooks Pate and co-workers,1−6 has become a standard instrument in molecular spectroscopy laboratories.7−16 Traditional rotational spectroscopy in the microwave and the millimeter-wave (mm-wave) regions is known for its exceptional resolution, which is ideal for precise determination of molecular structure.17 A rotational spectrum can also be used to unambiguously identify the species that are produced in a chemical reaction. A principal limitation of the pre-CP rotational spectrometer18,19 for chemical dynamics studies is that it is a narrowband apparatus. A scan over a wide spectral range that includes transitions from more than one species or conformer may require a prohibitively long time2 and does not ensure that the line intensities can be meaningfully compared. Broadband CP rotational spectroscopy opens new areas of chemical research while retaining the advantages of traditional pure rotational spectroscopy. Broad spectral regions are covered in a single chirped pulse, with meaningful relative line intensities in the resultant broadband spectra that can be converted to (vibrational state-specific) product branching ratios. CP rotational spectroscopy is a nearly universal (applicable to all polar species) chemical tool that is conformerand state-specific, quantitative, nondestructive, and suitable for studies of stable and transient species. Pate and co-workers have introduced the CP technique to isomerization dynamics studies by measuring the rate of isomerization between the syn© XXXX American Chemical Society

Received: August 20, 2014 Accepted: October 3, 2014

3641

dx.doi.org/10.1021/jz501758p | J. Phys. Chem. Lett. 2014, 5, 3641−3648

The Journal of Physical Chemistry Letters

Letter

without a barrier onto a PES plateau on the product side of the reaction until the radical products are completely separated. Roaming is a third scenario of unimolecular decomposition (iii), in which the system, upon reaching the above-mentioned plateau, reorients and “roams” over the PES while maintaining a few Å of separation between the two radical fragments. Eventually, a reactive configuration is found, at which the roaming radicals rearrange into two closed-shell molecular products. The roaming dynamics typically has its threshold in a relatively narrow (∼1 kcal/mol) energy window just below the SBF energy and is associated with a very loose roaming transition state (RTS). Transition state theory (TST) is commonly used to relate the properties of the TS to reaction rates. There have been proposals23 to extend TST to include roaming dynamics, and several approaches are in development,22−24 but this is a very challenging task. To benchmark these approaches, one needs an experimental method to measure the branching between these three reactive channels, but to date, even for very small systems, branching measurements between these three reactive pathways have not been achieved. The capability of CP spectroscopy to measure the branching into multiple products and, through that, the related reaction rates may prove to be a decisive technique for revealing roaming reaction mechanisms. The vibrational population distribution of photolysis12 products is also measurable in a CP experiment and encodes the properties of a TS. However, when studying complex chemistry involving secondary and higher-order reactions that follow the initial unimolecular decomposition event, kinetic modeling will also be essential to extract the reaction rates from the measured product concentrations. In this Letter, we use CP spectroscopy to detect and quantify three reaction products and, by means of kinetic simulation, show that the observed yield of nitrosyl hydride (HNO) relative to other products in the pyrolysis of ethyl nitrite (CH3CH2ONO) is evidence for roaming dynamics. The other two products that we measure are formaldehyde (CH2O) and acetaldehyde (CH3CHO). A similar system, methyl nitrite (CH3ONO), was studied in a recent theoretical work, and a roaming mechanism for CH3ONO → CH2O + HNO molecular elimination was suggested.34 We employ the Reaction Mechanism Generator (RMG) software that has been demonstrated to be a reliable tool for modeling complex chemical reactions.35 Introducing the roaming mechanisms into the model can affect, in a major way, the predicted rates of some reactions and thus the concentrations of products predicted by the model. We find a unique set of reactions and rates that reproduces the measured branching between the HNO, CH2O, and CH3CHO products of CH3CH2ONO pyrolysis and through that learn about the underlying dynamics and estimate the contributions of various mechanisms in the thermal decomposition of ethyl nitrite. The experimental setup (Figure 1), consisting of the CP millimeter-wave (CPmmW) spectrometer8 coupled to a Chen type nozzle,36−39 was discussed in detail in a recent publication,14 and some relevant aspects are reviewed in the Experimental Methods section. The precursor ethyl nitrite molecules, entrained in argon gas at 0.1% by mole, are introduced into the reactor by a pulsed valve that operates at the stagnation pressure of p0 = 3 bar. After passing through the heated SiC tube, in which the pyrolysis reactions take place, the products and remaining precursor molecules expand supersonically into the vacuum chamber and can be measured by the

Figure 1. Experimental setup schematic. The molecules exiting the heated tubular reactor expand supersonically into vacuum and are probed by the mm-wave beam of the CPmmW spectrometer. The interaction region is situated in the collision-free zone of the expansion and is represented as a cloud of polarized molecules. The mm-wave beam is broadcast by a horn antenna and is focused into the interaction region by a Teflon lens. The frequency of the mm-wave beam is swept linearly in time (chirped), as illustrated schematically by the wave in the figure. The other lens and the horn collect the free induction decay (FID) emitted by the polarized molecular rotational transitions. The FID is digitized by the CPmmW spectrometer and Fourier-transformed to a frequency domain rotational spectrum that contains the transitions, including their relative intensities, of all species covered by the chirp.8,14 Reproduced from ref 14 by permission of the PCCP Owner Societies.

CPmmW spectrometer. The CPmmW spectra of the CH2O, CH3CHO, and HNO products of ethyl nitrite pyrolysis and the corresponding branching ratio were previously observed at the reactor wall temperature of Twall = 1500 K.14 In order to investigate the chemistry inside the reactor, we measured the branching ratios of the three products of CH3CH2ONO pyrolysis at various reactor temperatures. Figure 2 shows the relative concentrations of CH2O, CH3CHO, and HNO measured by the CPmmW spectroscopy at Twall = 1000, 1200, 1500, and 1800 K. In the present work, we are only concerned with relative concentrations of products or branching ratios and leave analysis of the absolute concentrations for future studies.

Figure 2. Branching of the ethyl nitrite pyrolysis products determined in the CPmmW experiment. The concentrations of CH2O (black squares), CH3CHO (red circles), and HNO (blue triangles), measured at the reactor temperatures Twall of 1000, 1200, 1500, and 1800 K, are displayed with the experimental uncertainties. 3642

dx.doi.org/10.1021/jz501758p | J. Phys. Chem. Lett. 2014, 5, 3641−3648

The Journal of Physical Chemistry Letters

Letter

Figure 3. RMG-Py simulations at temperatures (a) T = 1000 and (b) 1500 K and p = 0.5 bar: CH2O, CH3CHO, and HNO mole fraction (product concentration) distribution as a function of time, t, for different values of α and reactor temperatures. Plateau regions of the mole fraction profiles do not change qualitatively at other pressures in the 0.1−1 bar range (see SI, Part 3). The dotted vertical lines in each panel indicate the 100 μs time point at which the molecules exit the reactor in our experiment. The intersections of the CH2O, CH3CHO, and HNO traces with the 100 μs line indicate the mole fractions relevant to this experiment.

kinetics in the flash pyrolysis reactor, the Python version of the Reaction Mechanism Generator (RMG-Py) software35 has been used to explore the complete reaction network. The automated procedure implemented in RMG-Py for kinetic model generation is designed to avoid the process of constructing kinetic models manually, which can be errorprone and can often lead to wrong conclusions. The principles of RMG-Py modeling and the details of the present simulation are described in the Computational Methods section. As output, RMG-Py automatically generates all possible reaction pathways and constructs a comprehensive (CHEMKIN43-format) list of relevant elementary reaction steps, including values of reaction rate constants and thermochemical properties. CHEMKIN simulations were performed in an isobaric homogeneous batch reactor model. We appreciate that such a model is an oversimplification of the real reactor used in the present experiment. The sufficiency of this model to the goal of understanding the role of roaming dynamics in the present system is discussed below. Computational fluid dynamics (CFD) simulations of continuous flow through a similar SiC tubular reactor39 and measurements in a different type of pyrolysis reactor44 show a pressure drop toward the end of the tube and temperature variations. We found that at lower temperatures, the product distribution profiles do not change within the 0.1−1 bar pressure range (Figures S2 and S3 in the SI) and use the values of the average pressure, p = 0.5 bar, and the temperature, T = Twall, inside the reactor. The list of the key chemical reactions that generate and consume the three observed products, CH2O, CH3CHO, and HNO, are summarized in Part 1 of the SI. The CHEMKIN file containing all of the reactions considered in the model generated by RMG-Py is supplied in the SI. RMG-Py confirms the key role of all reactions R1−R7 indicated above. However, the roaming unimolecular dissociation (reaction R7) requires special treatment. Roaming paths have not been a part of the RMG-Py reaction family set, and the rate of reaction R7 is not well-known. The nature of this reaction, as discussed above, is intimately related to the radical elimination channel (reaction R1). Therefore, in order to introduce the roaming mechanism

The residence time of the molecules inside the reactor is estimated39 to be ∼100 μs, which, as discussed below and in the Supporting Information (SI), is sufficient, at elevated temperatures, for both the initial unimolecular decomposition of ethyl nitrite and the subsequent chemistry to take place. It was proposed in the earlier ethyl nitrite pyrolysis studies37,40−42 that formaldehyde and acetaldehyde products can result from a SBF radical elimination reaction SBF

CH3CH 2ONO ⎯⎯⎯→ CH3CH 2O + NO

(R1)

followed by fast dissociation of the ethoxy radical (CH3CH2O) via either of two channels CH3CH 2O → CH 2O + CH3

(R2)

CH3CH 2O → CH3CHO + H

(R3)

or The mechanism leading to the formation of HNO is less certain. It may involve abstraction of H-atoms by free NO radicals CH3CH 2O + NO → CH3CHO + HNO

(R4)

RH + NO → R + HNO

(R5)

conventional unimolecular decomposition via a tight TS tight TS

CH3CH 2ONO ⎯⎯⎯⎯⎯⎯⎯→ CH3CHO + HNO

(R6)

or the NO radical roaming pathway RTS

CH3CH 2ONO ⎯⎯⎯→ CH3CHO + HNO

(R7)

Reaction R7 implies that after the initial cleavage of the CH3CH2O−NO bond, CH3CH2O and NO radicals roam in the vicinity of each other until NO abstracts an H-atom from its counterpart and the final CH3CHO and HNO reaction products are formed. The mechanisms associated with reactions R6, R1, and R7 are discussed above in (i), (ii), and (iii), respectively. While reactions R1−R6 are commonly accepted, additional, unknown reaction pathways may be responsible for the observed product yields. In order to correctly describe the 3643

dx.doi.org/10.1021/jz501758p | J. Phys. Chem. Lett. 2014, 5, 3641−3648

The Journal of Physical Chemistry Letters

Letter

products (10−8 ≤ t ≤ 10−7 s in Figure 3a) by a set of unimolecular decompositions (reactions R1, R2, R3, and R7) and remains steady during the subsequent plateaus of concentration (10−7 ≤ t ≤ 10−2 s in Figure 3a), with some perturbation due to product consumption. Because the branching ratio is formed very quickly and remains nearly constant for extended periods of time, the current model is applicable even if the effective residency time trxn during which the reactions occur is different from 100 μs. Because similar branching ratios are observed for all reactor pressures considered (Figure S3 in the SI), this suggests the results of the current model, which assumes a constant pressure throughout the reactor, would be immune to the effects of pressure dropping along the reactor axis that are predicted48 by the CFD calculations. We estimate that as long as the effective conditions in the reactor when Twall = 1000 K are within the 800 ≤ T ≤ 1000 K, 0.1 ≤ p ≤ 1 bar, 1 ≤ trxn ≤ 1000 μs volume of the (T, p, trxn) parameter space, the measured CH2O : CH3CHO : HNO branching ratio is described by the kinetic model in which most of the HNO product is formed by NO radical roaming in the thermal decomposition of CH3CH2ONO. Once the conditions in the reactor can be controlled with more certainty, it would be interesting to experimentally investigate the subtle question34 of how the unimolecular roaming pathway depends on pressure. Collisions with argon atoms may quench the roaming NO radical and shift the product branching from reaction R7 to reaction R1. In the present study this effect is implicitly included in the α(T) parameter. At higher temperatures (T = 1500 K in Figure 3b and T = 1200−1800 K in the SI, Parts 2 and 3) inside the reactor, we obtain poorer agreement of the predicted CH2O and CH3CHO concentrations with experiment. Likely contributors to this discrepancy are the rate constants for the ethoxy decomposition used in the present model, specifically those responsible for the branching between reactions R2 and R3. Although the rates from Batt49,50 measured at 393−433 K give more satisfactory agreement with our experiment, we use the data from the compilation of Curran50,51 for consistency with the other reaction rates that are taken from refs 50 and 51. At elevated temperatures, it is also the case that smaller α, within the 0.1 < α < 1 range, are required to explain the measured HNO branching. This is in accordance with the calculated increase of roaming branching at lower, near-threshold energies in a microcanonical ensemble21−23,25 and the observed temperature behavior of roaming in shock tube experiments.31,32 An important consideration in these experiments is the poorly understood role of wall reactions in the SiC tube that are capable of skewing the product distribution. The possibility of wall reactions was considered in a previous investigation52 by isotope scrambling experiments, leading to the conclusion that wall reactions were not an important factor in that study. A new CFD investigation48 calculates the characteristic time of diffusion of acetaldehyde molecules to the wall of the reactor and concludes that it is less but comparable to the residence time. In principle, there is a possibility for NO radicals that are formed in reaction R1 to abstract an H atom that may reside on the wall and form the HNO product. To elucidate the possible contributions of nonunimolecular mechanisms of HNO formation, we conducted a set of additional measurements. NO gas was premixed with CH3OD (Sigma-Aldrich, 99% atom D) in argon with mole fractions of 0.1% NO + 0.1% CH3OD + 99.8% Ar and expanded through

to the kinetic simulation, we adopt the following simple model of proportionality between the rates of reactions R1 (radical dissociation channel) and R7 (roaming channel) RTS

α (T ) =

k(CH3CH 2ONO ⎯⎯⎯→ CH3CHO + HNO) SBF

k(CH3CH 2ONO ⎯⎯⎯→ CH3CH 2O + NO) (E1)

with α(T) as the proportionality coefficient, which is a function of temperature. All RMG-Py calculations were repeated for different values of α, and the results at T = 1000 and 1500 K are displayed in Figure 3 (the simulation results at other temperatures and pressures are available in the SI). Each panel shows the time evolution of the concentrations of CH2O, CH3CHO, and HNO at a certain temperature, T. The t = 10−4 s time point corresponds to the residence time in the reactor, and thus, the product branching ratio at t = 10−4 s can be compared to the experimentally measured branching ratio in Figure 2. The best agreement between the experiment and the model is achieved at lower reactor temperatures. For example, at T = 1000 K and α = 1, (Figure 3a) the branching ratio CH2O : CH3CHO : HNO is close to 17.4 : 5.4 : 4.3 measured at Twall = 1000 K (Figure 2). Decomposition of ethyl nitrite starts at t = 10−9 s, and by the t = 10−7 s time point, formation of CH2O, CH3CHO, and HNO and decomposition of CH3CH2ONO are complete. An important observation is that, according to our kinetic model, bimolecular chemistry that sets in at about t = 10−5 s consumes the three products and does not produce HNO. Therefore, the observed HNO product is formed in a fast unimolecular decomposition reaction via a roaming path (reaction R7) or through a tight TS45 (reaction R6). Roaming is expected to have a much larger reaction rate because of both the entropic46,47 and energetic25,34 factors. Proceeding via a loose RTS, a system samples a much greater configuration space than that when dissociating through a conventional tight TS; in addition, the latest calculations34 of the CH3NO2/ CH3ONO system place the RTS in the CH3ONO → CH2O + HNO reaction 6.8 kcal/mol lower in energy than the corresponding tight TS. Indeed, in Figure 3a, when the parameter α is set to smaller values, effectively “switching off” the roaming mechanism, the predicted concentration of HNO at t = 10−4 s decreases to a level far below what is observed experimentally. The roaming mechanism (reaction R7) that is introduced into the kinetic model via equation E1 is required with α ∼ 1, for agreement between theory and experiment. Lin and co-workers, in their recent calculations of thermal decomposition of methyl nitrite,34 came to a similar conclusion: the SBF and the roaming channels are of equal importance for HNO formation (see Figure 7 in their work). The temperature distribution inside the reactor is likely to be nonuniform, with the regions adjacent to the reactor wall heated up to Twall and with colder core flow at T < Twall.14,48 In order to account for possible contributions of colder layers, the T = 600 and 800 K simulations are provided in the SI. Whereas no chemistry takes place before t ∼ 1000 μs at T = 600 K, roaming with α > 1 is required for modeling the T = 800 K flow layer inside the tube heated to Twall = 1000 K. Therefore, even (and especially) if HNO is formed in the colder regions of the reactor at T < Twall, it signifies the roaming dynamics. The RMG model leads to an important understanding of the kinetics at hand. The CH2O : CH3CHO : HNO branching ratio is defined during the initial exponentially fast formation of 3644

dx.doi.org/10.1021/jz501758p | J. Phys. Chem. Lett. 2014, 5, 3641−3648

The Journal of Physical Chemistry Letters

Letter

calculated ab initio to improve the agreement between the measured and predicted CH2O : CH3CHO branching ratios. Understanding the role of roaming in thermally induced reactions is important for combustion and biomass pyrolysis research and applications. Roaming is now known to be a ubiquitous phenomenon with sizable branching that needs to be considered in accurate kinetic modeling along with the conventional molecular and radical elimination mechanisms. A further investigation of roaming dynamics in the pyrolysis of ethyl nitrite as well as methyl nitrite and nitromethane would be possible and desirable. Detecting and quantifying additional pyrolysis products such as CO and NO would further facilitate gauging kinetic models against the CP measurements. Other chemical systems that may be studied using the combination of CP spectroscopy and kinetic modeling are pyrolysis of esters and carboxylic acids that are of importance for combustion research. Investigations of roaming in acetaldehyde, propanal and acetone are also within reach.

Figure 4. On the possibility of wall reactions. Yield of HNO in the pyrolysis of 0.1% CH3CH2ONO + 99.9% argon (squares) and in the pyrolysis of 0.1% NO + 0.1% CH3OD + 99.8% argon (circles) measured at the reactor temperatures Twall of 1000, 1200, 1500, and 1800 K. Displayed with the experimental uncertainties.

■

EXPERIMENTAL METHODS The Chen type flash pyrolysis reactor36 used in the present work is similar to the reactor used in the Ellison research group.38,39 The reactor is a 5.7 cm long, 1 mm i.d. SiC tube mounted inside the vacuum chamber that operates at the average pressure of 2.5 × 10−5 mbar. The 3.2 cm long section of the tube can be resistively heated to wall temperatures of Twall = 1000−1800 K as measured by an optical pyrometer with the experimental uncertainty of ±100 K. Despite the high temperature of the SiC reactor, the molecules are efficiently cooled in the expansion to rotational temperatures of Trot = 4 ± 1 K14 and can be studied by rotational spectroscopy. While rotational relaxation is considered to be fast for most species, vibrational relaxation is highly dependent on the molecule’s vibrational energy level separations, symmetry, and couplings.14 Unlike in the slit jet experiments12 where nascent vibrational population distributions are available for investigation, heavy rotation−vibration cooling in the expansion from the flash pyrolysis reactor, at least in argon carrier gas, imprints on the vibrational level population. Scrambling of the nascent vibrational population information is compensated for by the efficient rotational cooling in the present flash pyrolysis experiments, which facilitates rotational thermalization of reaction products and allows for product branching to be quantified. To maximize the overlap between the mm-wave beam and the molecular sample in the collision-free zone of the expansion (Figure 1), where the molecules reach their terminal rotational temperature Trot, no skimmer is used in this experiment. A CPmmW spectrometer8,14 that operates in the 63−102 GHz spectral region is utilized in this work in the multichirp14 mode of operation. Preference is given to the mm-wave spectral region4,6,8 over the more commonly used microwave region (2−50 GHz) in order to observe the more intense, at Trot = 4 ± 1 K, JKaKc = 101−000 rotational transitions53,54 of small (two heavy atoms) molecules, such as CH2O and HNO, and also higher-J transitions of larger molecules. Conversion of the experimentally observed rotational line intensities to product concentrations is performed14 using the PGOPHER54 software, which is set to include the rotational levels’ population differences in the intensity calculations. A rotational temperature of Trot = 4 ± 1 K measured for the acetaldehyde product molecules is assumed to be common to all species in the interaction region.14 The other required

the reactor at various temperatures. The observed yield of HNO is plotted in Figure 4 and compared to the results of ethyl nitrite pyrolysis. A DNO signal was not observed. Direct H-atom abstraction CH3OD + NO → CH2OD + HNO is expected to be slow, and the observed HNO is likely to be associated with the wall reactions. Thus, the NO and CH3OD mixture measurements suggest that wall reactions can be responsible for ≲10% of HNO formed in the ethyl nitrite pyrolysis experiments described in this work. Although the wallmediated reactions are likely to be present, they are expected to make only a minor contribution that is on par with other uncertainties in the present study. In summary, we suggest that the roaming mechanism plays an important role in the thermal decomposition of ethyl nitrite. RTS

The CH 3 CH 2 ONO ⎯⎯⎯→ CH 3 CHO + HNO reaction, proceeding via NO radical roaming, is found to be responsible for the appearance of most of the HNO product. We estimate that at the reactor temperature of Twall = 1000 K, the rate of RTS

decomposition by roaming, CH3CH2ONO ⎯⎯⎯→ CH3CHO + HNO, is of the same order of magnitude as the rate of SBF SBF

reaction CH3CH2ONO ⎯⎯⎯→ CH3CH2O + NO. Other reactions (unimolecular or bimolecular) are found not to make a significant contribution to the thermal decomposition of ethyl nitrite. Quantitative and nearly universal detection of multiple reaction products by CP Fourier transform rotational spectroscopy in combination with accurate kinetic modeling has shown to be a fruitful approach to studying molecular dynamics and kinetics of complex systems. The development of the CP method to the study of pyrolysis reactions in a Chen type nozzle will benefit from several improvements. The significance, and mechanisms, of wall reactions need to be clarified in more detail. Applying calculated48 or measured temperature and pressure fields in RMG-Py and CHEMKIN simulations will be a further major improvement in the accuracy of the modeling part of the present type of studies. Establishing more subtle properties of roaming dynamics through the exact value of the α parameter and its temperature and pressure dependencies would then be within reach. The rates of the crucial reactions, such as reactions R2 and R3 in the present work, can be measured or 3645

dx.doi.org/10.1021/jz501758p | J. Phys. Chem. Lett. 2014, 5, 3641−3648

The Journal of Physical Chemistry Letters

Letter

parameters are permanent electric dipole moments53 that are converted to transition electric dipole moments17 and the chirp bandwidths and their duration.

The CHEMKIN file is supplied. This material is available free of charge via the Internet at http://pubs.acs.org.

COMPUTATIONAL METHODS RMG-Py is a rate-rule-based reaction network generator that constructs reaction networks at given conditions (temperature, pressure, initial concentration) and user-defined tolerances for accuracy. RMG-Py estimates the high-pressure kinetics using over 40 reaction family templates filled with representative kinetics (e.g., H-abstraction, radical recombination, β-scission), including recently added nitrogen chemistry. This is supported by vast libraries of measured and calculated rates. RMG-Py constructs reaction mechanisms in a hierarchical manner. Starting from a seed mechanism, it continues extending the mechanism by searching for new reactions until all newly found reactions are slower than a predefined fraction of the characteristic rate constant of the overall mechanism. A seed mechanism is defined as a list of species, reactions, and rate coefficients that will remain part of the reaction mechanism and contains mostly small molecule reactions that are not well captured by characteristic rates defined in reaction families, which apply better for larger molecules, but that are essential for phenomena like ignition. The RMG-Py software then proposes new reactions, using the list of species supplied in the seed mechanism and the new larger molecules of interest, by comparing each combinatorial possibility against the reaction family templates in the RMG-Py database. A complete and continuously expanding list of reaction family templates can be found online.35 Any reactions composed exclusively of those species in the seed mechanism are added to the core of the reaction mechanism (the “cross-reactions”); all others are added to the edge of the reaction mechanism. Application of the rate-based algorithm55 for the temperature, pressure, and species’ concentrations of interest, until the user-specified tolerance is satisfied, results in the final chemical kinetic model. In the present case, the seed mechanism consists of smallmolecule chemistry that originates from the EFRC Foundation Fuel Mechanism v.0.956 and several nitrogen chemistry mechanisms.50,57−60 The simulation was run at pressures of 0.1, 0.5, and 1 bar and temperatures of T = 600, 800, 1000, 1200, 1500, and 1800 K to include low- and high-temperature chemistry in the model and for initial mole fractions of 0.001 for ethyl nitrite and 0.999 for Ar to mimic the conditions of the experiment. Pressure-dependent kinetics has been estimated by the Modified Strong Collision approach,61 using estimates of ρ(E) described by Allen et al.62 The thermochemistry of all species is either supplied by thermochemical libraries or estimated using Benson’s group additivity scheme.63 Transport properties, such as Lennard-Jones parameters, are estimated using an empirical relationship that depends on a species’ critical temperature and pressure.64 The termination criterion was 0.99% conversion of the initial reactants in all cases.

Corresponding Author

■

■

■

AUTHOR INFORMATION

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank Prof. Barney Ellison for his assistance in the development of the MIT pyrolysis experiment. K.P. acknowledges the Donors of the American Chemical Society Petroleum Research Fund for support under the Grant Number 50650ND6. R.W.F. and K.P. acknowledge the personnel and equipment support by the U.S. Department of Energy, Office of Basic Energy Sciences under Award Number DEFG0287ER13671. Y.V.S. and W.H.G. acknowledge the support of the Combustion Energy Frontier Research Center, an Energy Frontier Research Center funded by the U.S. Department of Energy, Office of Basic Energy Sciences under Award Number DE-SC0001198. A.G.S. acknowledges the Army Research Office Award Number 58245-CH-11 for financial support. B.B. acknowledges gratefully the financial support from the Swiss National Science Foundation (SNF) Postdoctoral Research Grant PBEZB2-140081.

■

REFERENCES

(1) Brown, G. G.; Dian, B. C.; Douglass, K. O.; Geyer, S. M.; Shipman, S. T.; Pate, B. H. A Broadband Fourier Transform Microwave Spectrometer Based on Chirped Pulse Excitation. Rev. Sci. Instrum. 2008, 79, 053103. (2) Dian, B. C.; Brown, G. G.; Douglass, K. O.; Pate, B. H. Measuring Picosecond Isomerization Kinetics via Broadband Microwave Spectroscopy. Science 2008, 320, 924−928. (3) Neill, J. L.; Douglass, K. O.; Pate, B. H.; Pratt, D. W. Next Generation Techniques in the High Resolution Spectroscopy of Biologically Relevant Molecules. Phys. Chem. Chem. Phys. 2011, 13, 7253−7262. (4) Steber, A. L.; Harris, B. J.; Neill, J. L.; Pate, B. H. An Arbitrary Waveform Generator Based Chirped Pulse Fourier Transform Spectrometer Operating from 260 to 295 GHz. J. Mol. Spectrosc. 2012, 280, 3−10. (5) Zaleski, D. P.; Neill, J. L.; Muckle, M. T.; Seifert, N. A.; Carroll, P. B.; Weaver, S. L. W.; Pate, B. H. A K-a-Band Chirped-Pulse Fourier Transform Microwave Spectrometer. J. Mol. Spectrosc. 2012, 280, 68− 76. (6) Neill, J. L.; Harris, B. J.; Steber, A. L.; Douglass, K. O.; Plusquellic, D. F.; Pate, B. H. Segmented Chirped-Pulse Fourier Transform Submillimeter Spectroscopy for Broadband Gas Analysis. Opt. Express 2013, 21, 19743−19749. (7) Karunatilaka, C.; Shirar, A. J.; Storck, G. L.; Hotopp, K. M.; Biddle, E. B.; Crawley, R.; Dian, B. C. Dissociation Pathways of 2,3Dihydrofuran Measured by Chirped-Pulse Fourier Transform Microwave Spectroscopy. J. Phys. Chem. Lett. 2010, 1, 1547−1551. (8) Park, G. B.; Steeves, A. H.; Kuyanov-Prozument, K.; Neill, J. L.; Field, R. W. Design and Evaluation of a Pulsed-Jet Chirped-Pulse millimeter-Wave Spectrometer for the 70−102 GHz Region. J. Chem. Phys. 2011, 135, 024202. (9) Pate, B. H.; De Lucia, F. C. Broadband Molecular Rotational Spectroscopy Special Issue Introduction. J. Mol. Spectrosc. 2012, 280, 1−2. (10) Schnell, M. Broadband Rotational Spectroscopy for Molecular Structure and Dynamics Studies. Z. Phys. Chem. 2013, 227, 1−21. (11) Perez, C.; Lobsiger, S.; Seifert, N. A.; Zaleski, D. P.; Temelso, B.; Shields, G. C.; Kisiel, Z.; Pate, B. H. Broadband Fourier Transform

ASSOCIATED CONTENT

S Supporting Information *

The rates of some reactions that are used in the present RMGPy model are given in Part 1 of the Supporting Information. The results of the RMG-Py simulation at temperatures of T = 600, 800, 1000, 1200, 1500, and 1800 K and α = 0.01. 0.1, 0.5, and 1, and the pressure p = 0.5 bar are shown in Part 2. The RMG-Py modeling results at p = 0.1, 0.5, and 1 bar, T = 1000, 1200, 1500, and 1800 K, and α = 0.01, 1 are shown in Part 3. 3646

dx.doi.org/10.1021/jz501758p | J. Phys. Chem. Lett. 2014, 5, 3641−3648

The Journal of Physical Chemistry Letters

Letter

Rotational Spectroscopy for Structure Determination: The Water Heptamer. Chem. Phys. Lett. 2013, 571, 1−15. (12) Prozument, K.; Shaver, R. G.; Ciuba, M. A.; Muenter, J. S.; Park, G. B.; Stanton, J. F.; Guo, H.; Wong, B. M.; Perry, D. S.; Field, R. W. A New Approach toward Transition State Spectroscopy. Faraday Discuss. 2013, 163, 33−57. (13) Seifert, N. A.; Zaleski, D. P.; Perez, C.; Neill, J. L.; Pate, B. H.; Vallejo-Lopez, M.; Lesarri, A.; Cocinero, E. J.; Castano, F.; Kleiner, I. Probing the C−H Weak Hydrogen Bond in Anesthetic Binding: The Sevoflurane−Benzene Cluster. Angew. Chem., Int. Ed. 2014, 53, 3210− 3213. (14) Prozument, K.; Park, G. B.; Shaver, G. R.; Vasiliou, A. K.; Oldham, J. M.; David, D. E.; Muenter, J. S.; Stanton, J. F.; Suits, A. G.; Ellison, G. B.; Field, R. W. Chirped-Pulse millimeter-Wave Spectroscopy for Dynamics and Kinetics Studies of Pyrolysis Reactions. Phys. Chem. Chem. Phys. 2014, 16, 15739−15751. (15) Patterson, D.; Schnell, M. New Studies on Molecular Chirality in the Gas Phase: Enantiomer Differentiation and Determination of Enantiomeric Excess. Phys. Chem. Chem. Phys. 2014, 16, 11114− 11123. (16) Kidwell, N. M.; Vaquero-Vara, V.; Ormond, T. K.; Buckingham, G. T.; Zhang, D.; Mehta-Hurt, D. N.; McCaslin, L.; Nimlos, M. R.; Daily, J. W.; Dian, B. C.; Stanton, J. F.; Ellison, G. B.; Zwier, T. S. Chirped-Pulse Fourier Transform Microwave Spectroscopy Coupled with a Flash Pyrolysis Microreactor: Structural Determination of the Reactive Intermediate Cyclopentadienone. J. Phys. Chem. Lett. 2014, 5, 2201−2207. (17) Townes, C. H.; Schawlow, A. L. Microwave Spectroscopy; McGraw Hill Book Company: New York, 1955. (18) Balle, T. J.; Flygare, W. H. Fabry−Perot Cavity Pulsed FourierTransform Microwave Spectrometer with a Pulsed Nozzle Particle Source. Rev. Sci. Instrum. 1981, 52, 33−45. (19) Suenram, R. D.; Grabow, J. U.; Zuban, A.; Leonov, I. A Portable, Pulsed-Molecular-Beam, Fourier-Transform Microwave Spectrometer Designed for Chemical Analysis. Rev. Sci. Instrum. 1999, 70, 2127− 2135. (20) Townsend, D.; Lahankar, S. A.; Lee, S. K.; Chambreau, S. D.; Suits, A. G.; Zhang, X.; Rheinecker, J.; Harding, L. B.; Bowman, J. M. The Roaming Atom: Straying from the Reaction Path in Formaldehyde Decomposition. Science 2004, 306, 1158−1161. (21) Zhang, X. B.; Rheinecker, J. L.; Bowman, J. M. Quasiclassical Trajectory Study of Formaldehyde Unimolecular Dissociation: H2CO → H2 + CO, H + HCO. J. Chem. Phys. 2005, 122, 114313. (22) Suits, A. G. Roaming Atoms and Radicals: A New Mechanism in Molecular Dissociation. Acc. Chem. Res. 2008, 41, 873−881. (23) Klippenstein, S. J.; Georgievskii, Y.; Harding, L. B. Statistical Theory for the Kinetics and Dynamics of Roaming Reactions. J. Phys. Chem. A 2011, 115, 14370−14381. (24) Bowman, J. M.; Shepler, B. C. Roaming Radicals. Annu. Rev. Phys. Chem. 2011, 62, 531−553. (25) Homayoon, Z.; Bowman, J. M. Quasiclassical Trajectory Study of CH3NO2 Decomposition via Roaming Mediated lsomerization Using a Global Potential Energy Surface. J. Phys. Chem. A 2013, 117, 11665−11672. (26) Bowman, J. M. Roaming. Mol. Phys. 2014, DOI: 10.1080/ 00268976.2014.897395. (27) Houston, P. L.; Kable, S. H. Photodissociation of Acetaldehyde as a Second Example of the Roaming Mechanism. Proc. Natl. Acad. U.S.A. 2006, 103, 16079−16082. (28) Grubb, M. P.; Warter, M. L.; Xiao, H. Y.; Maeda, S.; Morokuma, K.; North, S. W. No Straight Path: Roaming in Both Ground- and Excited-State Photolytic Channels of NO3 → NO + O2. Science 2012, 335, 1075−1078. (29) Dey, A.; Fernando, R.; Abeysekera, C.; Homayoon, Z.; Bowman, J. M.; Suits, A. G. Photodissociation Dynamics of Nitromethane and Methyl Nitrite by Infrared Multiphoton Dissociation Imaging with Quasiclassical Trajectory Calculations: Signatures of the Roaming Pathway. J. Chem. Phys. 2014, 140, 054305.

(30) Sivaramakrishnan, R.; Michael, J. V.; Klippenstein, S. J. Direct Observation of Roaming Radicals in the Thermal Decomposition of Acetaldehyde. J. Phys. Chem. A 2010, 114, 755−764. (31) Sivaramakrishnan, R.; Michael, J. V.; Wagner, A. F.; Dawes, R.; Jasper, A. W.; Harding, L. B.; Georgievskii, Y.; Klippenstein, S. J. Roaming Radicals in the Thermal Decomposition of Dimethyl Ether: Experiment and Theory. Combust. Flame 2011, 158, 618−632. (32) Sivaramakrishnan, R.; Michael, J. V.; Harding, L. B.; Klippenstein, S. J. Shock Tube Explorations of Roaming Radical Mechanisms: The Decompositions of Isobutane and Neopentane. J. Phys. Chem. A 2012, 116, 5981−5989. (33) Joalland, B.; Shi, Y.; Kamasah, A.; Suits, A. G.; Mebel, A. M. Roaming Dynamics in Radical Addition−Elimination Reactions. Nat. Commun. 2014, 5, 4064. (34) Zhu, R. S.; Raghunath, P.; Lin, M. C. Effect of Roaming Transition States upon Product Branching in the Thermal Decomposition of CH3NO2. J. Phys. Chem. A 2013, 117, 7308−7313. (35) Green, W. H.; Allen, J. W.; Buesser, B. A.; Ashcraft, R. W.; Beran, G. J.; Class, C. A.; Gao, C.; Goldsmith, C. F.; Harper, M. R.; Jalan, A.; Keceli, M.; Magoon, G. R.; Matheu, D. M.; Merchant, S. S.; Mo, J. D.; Petway, S.; Raman, S.; Sharma, S.; Song, J.; Suleymanov, Y. V.; Van Geem, K. M.; Vandeputte, A. G.; Wen, J.; West, R. H.; Wong, A.; Wong, H.-W.; Yelvington, P. E.; Yee, N.; Yu, J. RMG - Reaction Mechanism Generator. https://github.com/GreenGroup/RMG-Py (2014). (36) Kohn, D. W.; Clauberg, H.; Chen, P. Flash Pyrolysis Nozzle for Generation of Radicals in a Supersonic Jet Expansion. Rev. Sci. Instrum. 1992, 63, 4003−4005. (37) Rohrs, H. W.; Wickhamjones, C. T.; Ellison, G. B.; Berry, D.; Argrow, B. M. Fourier-Transform Infrared-Absorption Spectroscopy of Jet-Cooled Radicals. Rev. Sci. Instrum. 1995, 66, 2430−2441. (38) Zhang, X.; Friderichsen, A. V.; Nandi, S.; Ellison, G. B.; David, D. E.; McKinnon, J. T.; Lindeman, T. G.; Dayton, D. C.; Nimlos, M. R. Intense, Hyperthermal Source of Organic Radicals for MatrixIsolation Spectroscopy. Rev. Sci. Instrum. 2003, 74, 3077−3086. (39) Urness, K. N.; Guan, Q.; Golan, A.; Daily, J. W.; Nimlos, M. R.; Stanton, J. F.; Ahmed, M.; Ellison, G. B. Pyrolysis of Furan in a Microreactor. J. Chem. Phys. 2013, 139, 124305. (40) Steacie, E. W. R.; Shaw, G. T. The Homogeneous Unimolecular Decomposition of Gaseous Alkyl Nitrites. II. The Decomposition of Ethyl Nitrite. J. Chem. Phys. 1934, 2, 345. (41) Levy, J. B. The Thermal Decomposition of Ethyl Nitrite. J. Am. Chem. Soc. 1956, 78, 1780−1783. (42) Batt, L.; Milne, R. T. Gas-Phase Pyrolysis of Alkyl Nitrites. IV. Ethyl Nitrite. Int. J. Chem. Kinet. 1977, 9, 549−565. (43) CHEMKIN-MFC, MFC 5.0; Reaction Design: San Diego, CA, 2010. (44) Zhang, Y. J.; Cai, J. H.; Zhao, L.; Yang, J. Z.; Jin, H. F.; Cheng, Z. J.; Li, Y. Y.; Zhang, L. D.; Qi, F. An Experimental and Kinetic Modeling Study of Three Butene Isomers Pyrolysis at Low Pressure. Combust. Flame 2012, 159, 905−917. (45) Fernández-Ramos, A.; Martínez-Núñez, E.; Ríos, M. A.; Rodríguez-Otero, J.; Vázquez, S. A.; Estévez, C. M. Direct Dynamics Study of the Dissociation and Elimination Channels in the Thermal Decomposition of Methyl Nitrite. J. Am. Chem. Soc. 1998, 120, 7594− 7601. (46) Heazlewood, B. R.; Jordan, M. J. T.; Kable, S. H.; Selby, T. M.; Osborn, D. L.; Shepler, B. C.; Braams, B. J.; Bowman, J. M. Roaming is the Dominant Mechanism for Molecular Products in Acetaldehyde Photodissociation. Proc. Natl. Acad. U.S.A. 2008, 105, 12719−12724. (47) Harding, L. B.; Georgievskii, Y.; Klippenstein, S. J. Roaming Radical Kinetics in the Decomposition of Acetaldehyde. J. Phys. Chem. A 2010, 114, 765−777. (48) Guan, Q.; Urness, K. N.; Ormond, T. K.; David, D. E.; Ellison, G. B.; Daily, J. W. The Properties of a Micro-Reactor for the Study of the Unimolecular Decomposition of Large Molecules. Int. J. Phys. Chem. 2014, Submitted. (49) Batt, L. Gas-Phase Decomposition of Alkoxy Radicals. Int. J. Chem. Kinet. 1979, 11, 977−993. 3647

dx.doi.org/10.1021/jz501758p | J. Phys. Chem. Lett. 2014, 5, 3641−3648

The Journal of Physical Chemistry Letters

Letter

(50) Manion, J. A.; Huie, R. E.; Levin, R. D.; Burgess, D. R., Jr.; Orkin, V. L.; Tsang, W.; McGivern, W. S.; Hudgens, J. W.; Knyazev, V. D.; Atkinson, D. B.; Chai, E.; Tereza, A. M.; Lin, C.-Y.; Allison, T. C.; Mallard, W. G.; Westley, F.; Herron, J. T.; Hampson, R. F.; Frizzell, D. H. NIST Chemical Kinetics Database, NIST Standard Reference Database 17, version 7.0 (Web Version), release 1.4.3, data version 2008.12. http://kinetics.nist.gov (2008). (51) Curran, H. J. Rate Constant Estimation for C1 to C4 Alkyl and Alkoxyl Radical Decomposition. Int. J. Chem. Kinet. 2006, 38, 250− 275. (52) Vasiliou, A. K.; Piech, K. M.; Reed, B.; Zhang, X.; Nimlos, M. R.; Ahmed, M.; Golan, A.; Kostko, O.; Osborn, D. L.; David, D. E.; Urness, K. N.; Daily, J. W.; Stanton, J. F.; Ellison, G. B. Thermal Decomposition of CH3CHO Studied by Matrix Infrared Spectroscopy and Photoionization Mass Spectroscopy. J. Chem. Phys. 2012, 137, 164308. (53) Pickett, H. M.; Poynter, R. L.; Cohen, E. A.; Delitsky, M. L.; Pearson, J. C.; Muller, H. S. P. Submillimeter, Millimeter, and Microwave Spectral Line Catalog. J. Quant. Spectrosc. Radiat. Transfer 1998, 60, 883−890 ; see also: http://spec.jpl.nasa.gov. (54) Western, C. M. PGOPHER, a Program for Simulating Rotational Structure. University of Bristol: Bristol, U.K.; http://pgopher.chm.bris. ac.uk (2014). (55) Susnow, R. G.; Dean, A. M.; Green, W. H.; Peczak, P.; Broadbelt, L. J. Rate-Based Construction of Kinetic Models for Complex Systems. J. Phys. Chem. A 1997, 101, 3731−3740. (56) Wang, H. EFRC foundation kinetic, model v.0.9, Private Communication. (57) Dean, A. M.; Bozzelli, J. W. Combustion Chemistry of Nitrogen. In Gas-Phase Combustion Chemistry, Gardiner, W. C., Ed.; Springer: New York, 2000; pp 125−341. (58) Zhang, K. W.; Zhang, L. D.; Xie, M. F.; Ye, L. L.; Zhang, F.; Glarborg, P.; Qi, F. An Experimental and Kinetic Modeling Study of Premixed Nitroethane Flames at Low Pressure. Proc. Combust. Inst. 2013, 34, 617−624. (59) Lucassen, A.; Zhang, K. W.; Warkentin, J.; Moshammer, K.; Glarborg, P.; Marshall, P.; Kohse-Hoinghaus, K. Fuel−Nitrogen Conversion in the Combustion of Small Amines Using Dimethylamine and Ethylamine as Biomass-Related Model Fuels. Combust. Flame 2012, 159, 2254−2279. (60) Lopez, J. G.; Rasmussen, C. L.; Alzueta, M. U.; Gao, Y.; Marshall, P.; Glarborg, P. Experimental and Kinetic Modeling Study of C2H4 Oxidation at High Pressure. Proc. Combust. Inst. 2009, 32, 367− 375. (61) Chang, A. Y.; Bozzelli, J. W.; Dean, A. M. Kinetic Analysis of Complex Chemical Activation and Unimolecular Dissociation Reactions Using QRRK Theory and the Modified Strong Collision Approximation. Z. Chem. Phys. 2000, 214, 1533−1568. (62) Allen, J. W.; Goldsmith, C. F.; Green, W. H. Automatic Estimation of Pressure-Dependent Rate Coefficients. Phys. Chem. Chem. Phys. 2012, 14, 1131−1155. (63) Benson, S. W. Thermochemical Kinetics: Methods for the Estimation of Thermochemical Data and Rate Parameters; Wiley: New York, 1976 (64) Welty, J.; Wicks, C.; Wilson, R.; Rorrer, G. Fundamentals of Momentum, Heat, and Mass Transfer; Wiley: New York, 2000.

3648

dx.doi.org/10.1021/jz501758p | J. Phys. Chem. Lett. 2014, 5, 3641−3648